voice_client_local

#!/bin/bash
# Copyright (C) 2026 by Henry Kroll III, www.thenerdshow.com
#
# With code contributions by Matthew Rensberry
#
# This is free software.  You may redistribute it under the terms
# of the Apache license and the GNU General Public License Version
# 2 or at your option any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public
# License along with this program; if not, write to the Free Software
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
#

# fifo queue to hold temporary audio file names
audio_fifo=$(mktemp); rm "$audio_fifo" ; mkfifo "$audio_fifo"

# flag parsing
use_flow=false
if [[ "$1" == "-flow" ]]; then
    use_flow=true
fi

# start whisper service
systemctl --user start whisper.service
# tab away
# sleep 0.25 && ydotool key 56:0 42:0 56:1 15:1 56:0 15:0

## create a trap to remove temp files, stop server on untimely exit
cleanup() {
    systemctl --user stop whisper.service
    rm -f /tmp/tmp.txt "$audio_fifo"
}
trap cleanup 0

# function to process audio files from queue
trans(){
    while read audio; do
        # Send audio file to whisper.cpp using curl
        curl_output=$(curl -s "http://127.0.0.1:7777/inference" \
                        -H "Content-Type: multipart/form-data" \
                        -F "file=@$audio" \
                        -F "temperature=0.2" \
                        -F "response-format=json")

        # Parse JSON to extract text
        extracted_text=$(echo "$curl_output" | jq -r '.text')

        # remove temporary audio file
        rm -f "$audio"
        echo "$extracted_text"
        # Type text to terminal, in background
        # Thanks for watching! seems to occur frequently due to noise.
        if [[ ${#extracted_text} > 15 ]] && ! [[ "$extracted_text" =~ "Thanks for watching" ]] \
            && ! [[ "$extracted_text" =~ '[' ]]; then
            if [ "$use_flow" = true ]; then
                ydotool type $(llama_edit "$extracted_text")
            else
                ydotool type "$extracted_text"
            fi
        fi &
    done < "$audio_fifo"
    #cleanup
    rm -f "$audio_fifo"
}

# record audio in background
while true; do
    # Make temporary files to hold audio
    tmp=$(mktemp)
    # Remove it on exit
    trap 'rm "$tmp" ; exit' INT

    # Listen to the mic.
    # The `1 0.2 6%` part of this command trims one clip of silence
    # from the beginning. 6% of max volume is considered silence here.
    # `1 1.0 8%` stops after a 1.0 second pause in speech.
    # If recording never stops, increase sound threshold from 8%
    # to 10% or more. This can happen with poor recording equipment
    # or noisy environments with heaters and fans going.

    rec -c 1 -r 22050 -t mp3 "$tmp" silence 1 0.2 6% 1 1.0 8%

    # echo temporary audio file name to transcription queue
    echo "$tmp"
done > "$audio_fifo" & #The `&` lets recording run in the background.

# run audio transciption handler
trans